Minimally Invasive Sub-Retinal Transplantation of RPE-J Cells on a Biodegradable Composite PCL/Collagen Nanosheet

Abstract

Retinal cells are irreparably damaged by diseases such as age-related macular degeneration (AMD). A promising method to restore partial or whole vision is through cell-based transplantation to the damaged location. However, cell transplantation using conventional vitreous surgery is an invasive procedure that may induce infections and has a high failure rate of cell engraftment. In this study, we describe the fabrication of a biodegradable composite nanosheet used as a substrate to support retinal pigment epithelial (RPE-J) cells, which can be grafted to the sub-retinal space using a minimally invasive approach. The nanosheet was fabricated using polycaprolactone (PCL) and collagen in 80:20 weight ratio, and had size of 200 µm in diameter and 300 nm in thickness. These PCL/collagen nanosheets showed excellent biocompatibility and mechanical strength in vitro. Using a custom designed 27-gauge glass needle, we successfully transplanted an RPE-J cell loaded nanosheet into the sub-retinal space of a rat model with damaged photoreceptors. The cell loaded nanosheet did not trigger immunological reaction within 2 weeks of implantation and restored the retinal environment. Thus, this composite PCL/collagen nanosheet holds great promise for organized cell transplantation, and the treatment of retinal diseases.

Introduction

The view is a biological function that is directly related to people’s quality of life^1–3. Globally, in 2015, there were an estimated of 253 million visually impaired people, of which 36 million were blind ⁴ . In Japan, accordingly to a 2001 estimate, 1.64 million were visually impaired people, of which 187,000 were blind, and glaucoma, degenerative myopia, diabetic retinopathy, and age-related macular degeneration (AMD) accounted for roughly 57% of blindness ⁵ . The major cause of blindness is AMD, which is a disease related to retinal pigment epithelial (RPE) cells damages. The RPE cells have limited ability to regenerate in vivo; therefore, most treatment methods focus on slowing the disease progression through the use of drugs or cell secreted compounds^6,7. However, drugs are unable to fully restore RPE functions, which implies the urgent needs to develop reliable methods to fully treat damaged retina.

Stem cell technology advancements, such as the generation of induced pluripotent stem cells, allow regrowing RPE cells in vitro, promoting the use of cell-based therapy in the treatment of retinal degenerative diseases^8–10. The RPE is a pigmented monolayer of retinal cells that exists between the neural retina and the choroid, and which plays an important role in the maintenance and functions of the retina and its photoreceptors ¹¹ . This include maintaining the blood-retinal barrier, absorbing stray light, supplying nutrients to the neural retina, regenerating visual pigments, and absorbing and recycling shed photoreceptor outer segments ¹² . Additionally, RPE cells secreted brain-derived neurotrophic factor (BDNF) and pigment epithelium-derived factor (PEDF) which work together to maintain the homeostasis in blood vessel growth and suppression, respectively, in the retina^2,13. Using a tissue engineering approach, a successful transplantation of RPE cells should allow effective reconstruction of the extracellular matrix (ECM) and recovery of the photoreceptor functions ¹⁴ .

Engineering strategies allowing cell transport to the damaged region, and ensuring the cell survival in the body environment are necessary for a reliable cell transplant method. In this regard, growing cells into a cell sheet and then transferring them to the targeted site is a strategy that enables successful treatment^15–17. Compared to trypsin-induced detachment of cultured cells, cell sheet method using temperature sensitive substrates for cell cultures allows retaining adhesion factors, ECM, and cell layer organization, which ensures superior adhesion of cells to the host tissue, and expression of biomolecules enabling their long-term survival^18,19. However, cell delivery by conventional methods such as vitrectomy surgery requires multiple punctures of the eyeball to allow the transportation of the cell sheet to the tissue location through hollow needles ²⁰ . This delicate procedure is invasive with incisions of the sclera and retinal tissue, resulting in vitreous humor leakage and postoperative infections^21,22. Therefore, minimally invasive implantation with minimal damage to host tissue is critical to increase transplant success.

Cell injections have been used as a first approach for cell transplantation into different tissues ²³ . However, several difficulties appeared such as the poor cell survival rate at the site of transplantation, the low retention rate of transplanted cells, and the poor cell integration with the host with potential immune rejection ²⁴ . Thus, the transplantation of autologous RPE cells in AMD patients by injection of a cell suspension ²⁵ have been used, but limited visual improvement have been obtained due to low cell viability, low homogeneous cell distribution, and low integration into the subretinal tissue^26–28. The cell survival rate during and after transplantation by injection of a cell suspension is a major challenge. Indeed several factors contribute to the induction of cell death among which the mechanical force exerted on cells during injection ²⁹ . Thus, the velocity of a saline solution (Newtonian fluid) in the syringe is different at the center of the syringe and near its wall, which induces shear and extensional forces on cells. Furthermore, the difference in diameter between the syringe and the needle induces a sharp increase in shear and extensional forces that may affect the cell membrane integrity ³⁰ . Another problem with cell suspension is that many cell types need to adhere to a matrix to avoid anoikis^29,31. In addition, the transplanted cells do not find an optimized environment in the host degenerative or damaged tissue (e.g., oxidative stress, poor vascularization, disorganized matrix, and lack of growth factors) and their integration to the host tissue is challenging ³² . To overcome these problems, researchers have developed news approaches using biomaterials to support cells during the culture phase, to protect them during transplantation, and to ensure good engraftment of the transplanted cells^29,31. Thus, shear-thinning materials such as hyaluronan-methylcellulose hydrogel ³² or alginate ³⁰ have been used to protect cells during injection, whereas growth factors loaded hydrogels have been used to modify the microenvironment in the host tissue ³³ , and possibly to favor cell engraftment.

Among biomaterials that were used for RPE cell transplantation are ultrathin polymeric membranes or nanosheets. These nanosheets have unique interfacial and mechanical properties due to their huge size aspect ratios (between the size of the nanosheet surface and its thickness) and showed high flexibility, conformability, physical adhesiveness, and molecular permeability^27,34–36. They are particularly suitable for the delivery of organized cells in narrow spaces such as the sub-retinal space. Thus, our group has previously demonstrated successful transplantation of RPE-J cells, which are an immortalized RPE cell line from rat and kept numerous characteristics of RPE cells ³⁷ , using polylactic-co-glycolic acid (PLGA) ultrathin polymeric films of nanometer thickness as a substrate for cells, to overcome cell transplantation failure^27,38. The method used a 24 G (470 µm inner diameter) gauge needle and allowed the transplantation of cell loaded nanosheets in a minimally invasive way. In addition, we have extended our works to the fabrication of PLGA nanoribbons and showed their use in directing cell orientation and organization ³⁹ , whereas nanoribbons functionalized with 3,4-dihydroxy-l-phenylalanine (DOPA) to immobilize poly-l-lysine (PLL) and fibronectin allowed enhancing the cell adhesion and proliferation and could be injected via a 27 G (220 µm inner diameter) gauge needle ⁴⁰ . Compared to cell injection (without biomaterial), the use of biomaterial as micro/nano-substrates for cells allow protecting and supporting cells during injection due to the high flexibility of the nanosheets, the transplantation of organized cells, localizing the cells at the transplanted site, to enhance the cell retention, viability, and proliferation, and therefore to favor the repair of the tissue.

However, several studies have shown that when decomposed in vivo, PLGA scaffold produces polylactic acid and polyglycolic acid as byproducts, with acid dissociation constants in the lower end at 3.8, which increases the risk of inflammation post-transplantation^41,42. Moreover, although there is a consensus on the safety of PLGA microspheres used as a carrier for ocular drug delivery ⁴³ , some study reported that the toxicity of PLGA may be dependent of the size and shape of the PLGA carrier since PLGA microspheres (diameter 20–100 µm) tested in monkeys and rabbit eyes induced a strong immune response, while larger (0.9 mm x 3.7 mm) PLGA rods were well tolerated ⁴⁴ .

In this study, we report on the use of another type of nanosheet scaffold with excellent biocompatibility, cell supportive properties, tunable mechanical properties, excellent flexibility, and good mechanical strength (Fig. 1A). This time, we fabricated a composite nanosheet of PCL/ collagen, which has several advantages over PLGA nanosheet. First, PCL is a biodegradable synthetic polymer which is hydrolyzed under physiological conditions and produces caproic acid which has an acid dissociation constant around 4.9^45,46. Therefore, the risk of potential inflammation post-transplantation is less than for PLGA, since the PCL dissociation is effective at higher pH. Second, PCL degrades in vivo more slowly than PLGA making it suitable for long-term drug and growth factor delivery that can improve cell functionalities and cell integration to the host ⁴⁷ , whereas providing support and structural stability in the host tissue. Third, PCL-based nanosheets showed good flexibility with usually lower Young’s modulus than PLGA nanosheets, which gives to cells better protection during injection and potentially allows the use of needles with thinner diameter^45,48. Moreover, the presence of collagen, which is a natural polymer and a major constituent of the ECM, promotes the cell proliferation and functions, and enhances the cytocompatibility of PCL ⁴⁷ . Thus, RPE-J cells were grown on the composite nanosheet and a high cell viability (>80%) was obtained. The composite nanosheet showed excellent mechanical strength, could be easily handled, and passed through a syringe needle without breaking. In order to transplant cells into the delicate tissue of eye, we used a custom designed 27 G glass needle coated with 2-methacryloyloxyethyl phosphorylcholine (MPC) and were able to successfully deliver the cells in the sub-retinal space through the sclera of the rat in a minimally invasive and effective manner, and no inflammation was observed post-surgery (Fig. 1B). Therefore, this sub-retinal implant of cell loaded PCL/collagen nanosheet is a promising strategy for cell monolayer transplantation, and for the treatment of retinal degenerative diseases.

Figure 1.

Overview of the transplantation of RPE cells on a nanosheet in the subretinal space. (A) RPE cells were cultured on a composite nanosheet made of PCL/collagen. (B) Schematic showing RPE cells transplanted into the subretinal space through the sclera with a 27-gauge glass needle. RPE: retinal pigment epithelium; PCL: polycaprolactone.

Materials and Methods

Fabrication of PLGA, PCL, and PCL/Collagen Nanosheets

Spin coating and microcontact printing processes were combined to fabricate three different types of nanosheets made of PLGA (75:25, Mw: 97000, Polyscience, Inc, Warrington, PA, USA), PCL (average Mn 45,000, Sigma-Aldrich, Munich, Germany), and composite of PCL/collagen (Collagen Type I, powder, pig skin, pepsin-solubilized, NIP, Tokyo, Japan). Figure 2 is a schematic showing the fabrication process of PCL/collagen nanosheets. Two photos were added to show the polydimethylsiloxane (PDMS) stamp used in the process and the PCL/collagen nanosheet obtained at the end of the process. Thus, arrays of PDMS pillars with 200 µm diameter and 100 µm height were fabricated by curing PDMS (SYLGARDTM 184 Silicone Elastomer Kit, Dow Toray Co., Ltd, Tokyo, Japan) with curing agent (ratio 10:1) at 80^oC for 4 h on an epoxy resin (SU-8 2100) mold fabricated using soft lithography process ⁴⁹ . To prepare the polymer solutions, PLGA and PCL were dissolved in dichloromethane (Wako Pure Chemical, Osaka, Japan) at a final concentration of 5 mg/ml and 20 mg/ml, respectively. Since dichloromethane did not readily dissolve collagen, the mixture of PCL and collagen (80:20 weight ratio) was dissolved in 1,1,1,3,3,3-hexafluoro-2-propanol (HFP, Sigma-Aldrich, Munich, Germany) at a final concentration of 20 mg/ml. Polyvinyl alcohol (PVA, Mw 145,000, Sigma-Aldrich, Munich, Germany) was dissolved in water at 10 mg/ml. The polymer mixture of PCL/collagen was combined with Nile Red dye (Wako Pure Chemical, Osaka, Japan) for better visualization, which was added at 0.2 mg/ml. PVA was spin coated on a glass slide, whereas PLGA, PCL, and PCL/collagen were spin coated on a PDMS stamp for 40 s at 4000 rpm. The PDMS stamps were then imprinted on the PVA-coated glasses, which were then baked at 120^oC for 90 s and 10 min, respectively, for PLGA and PCL/collagen coated PDMS. Finally, by peeling the PDMS stamp off the PVA coated glass substrate, patterned nanosheets were obtained.

Figure 2.

Overview of the fabrication of composite PCL/collagen nanosheets. PDMS stamp was spin coated with PCL/collagen solution, whereas a glass plate was spin coated with PVA. The PDMS stamp was then stamped on the PVA glass plate, which resulted in imprints of PCL/collagen nanosheets on the PVA base. The inserts show images of the PDMS stamps and PCL/Collagen nanosheet. PCL: polycaprolactone; PDMS: polydimethylsiloxane; PVA: polyvinyl alcohol.

Tensile Test of PLGA, PCL, and Composite PCL/Collagen Nanosheets

A vertical electric test stand (EMX-1000N, Imada, Toyohashi, Japan) and a standard type digital force gauge were used to determine the mechanical properties of nanosheets (ZTS-200N, Imada, Toyohashi, Japan). However, to overcome the problems of gripping and fixing a nanosheet on the machine, we fabricated larger polymeric sheet samples with dimensions of 20 mm, 8 mm, and 10 µm, and used them for tensile test. PCL, and PCL/collagen nanosheets were made by first dissolving the polymers in HFP and then evaporating the solvents from a silicon mold. A uniaxial tensile test was done, and the resulting data were used to construct a stress–strain diagram.

Thickness of PCL/Collagen Nanosheet

The PCL/collagen nanosheets were fabricated by adjusting the spin coating speed and the solution concentrations. First, a final concentration of 20 mg/ml of PCL/collagen in an 80:20 weight ratio was spin coated for 40 s at varying speeds ranging from 4,000 to 8,000 rpm. Second, we changed the final concentration of PCL/collagen mixture, and spin coated at constant speed of 4,000 rpm for 40 s, and the resulting thickness was assessed.

Biodegradability of PLGA, PCL, and Composite PCL/Collagen Nanosheets

PLGA, PCL, and PCL/collagen solutions were spin coated on a glass surface for 40 s at 4,000 rpm. The solutions were evaporated for 10 min at 120^°C. Lipase (from Pseudomonas cepacian, light beige powder, 30 U/mg, Sigma-Aldrich, Munich, Germany) was dissolved in phosphate buffered saline (PBS) at 0.1 mg/ml. Samples of polymers were cut and placed in a 6-well plate with 4 ml lipase solution, and the plate was shaken on a microplate mixer at 1,500 rpm (NS-4P, AS ONE, Osaka, Japan). Samples were taken, dried, and weighed regularly over 200 days to determine the mass loss.

Water Contact Angle Measurement

In order to measure the radius and height of the droplets, 15 µl of distilled water was dropped on glass substrates spin-coated with PLGA, PCL, or PCL/collagen after drying. The glass surfaces were then photographed using a stereoscopic microscope (Digital Microscope KH-1300, Hirox, Tokyo, Japan).

RPE-J Cell Culture

RPE-J were obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA) and cultured at 33°C and 5% CO₂ in growth medium (Dulbecco’s modified Eagle medium (DMEM) containing 4% (v/v) fetal bovine serum (FBS) and 1% (v/v) antibiotic-antimycotic). When the cells reached around 70% confluence, they were detached from the culture dish using 0.25% (w/v) trypsin and 0.1% (w/v) ethylenediaminetetraacetic acid (EDTA), and the cell suspension was then sub-cultured or used in experiments.

Cell Culture on Nanosheets

The nanosheets were rinsed with Dulbecco’s phosphate-buffered saline (DPBS (-)), treated with oxygen plasma using a plasma generator apparatus (PM100, Yamato Kagaku, Tokyo, Japan), and exposed to UV in a sterilizing storage cabinet (NB-5 type, Nisshin Kogyo, Toyama, Japan). Cells were seeded at 4 × 10⁴ cells/cm² on the nanosheets. After 3 h incubation at 33°C and 5% CO₂, the culture medium was changed and the culture started.

Cellular Viability

Calcein-AM stock solution (1 mM, Dojindo, Kumamoto, Japan) and propidium iodide (PI) stock solution (1.5 mM, Dojindo, Kumamoto, Japan) were used to stain live (green fluorescence) and dead (red fluorescence) cells, respectively, following manufacturer’s protocol. Briefly, the cells on nanosheets in 6-well plate were rinsed with DPBS (-). Calcein-AM was adjusted to 2 µM and PI was adjusted to 4 µM in DPBS (-), then Calcein-AM/PI solution were added to each well and incubated for 15 min. Cells were counted and fluorescence images were taken for further image analysis.

Immunofluorescent Staining

Cells were immunostained to visualize tight junctions. Thus, cells were rinsed with DPBS(-) and fixed in 1.6% paraformaldehyde for 12 min. Then, they were permeabilized with Triton X-100 in DPBS(-) at a concentration of 3 µl/ml for 10 min, and blocked with 10 µl/ml bovine serum albumin (BSA) in DPBS(-) for 20 min at 37°C. They were incubated at 4°C for 24 h with the primary antibody rabbit anti-ZO-1(N-term) at 10 µl/ml in DPBS(-) with 0.1% BSA. After DPBS(-) rinses, cells were stained with the secondary antibody Alexa-Fluor 546-conjugated donkey anti-rabbit IgG in DPBS(-) at 2 µl/ml with 0.1% BSA, and incubated for 1 h at 37°C. They were also stained with 4’,6-diamidino-2-phenylindole (DAPI) 20 µl/ml in DPBS(-) for 15 min at 37°C.

Image Analysis and Cell Morphology Evaluation

To assess cell morphology, Voronoi division was used. Images of cells on nanosheets were taken, and the position of the nucleus centroid was determined using image processing software (Image J). A Voronoi diagram was created after performing Voronoi division based on the position of the center of gravity.

Fabrication of Glass Capillary Needle

We fabricated a glass needle delivery system for minimally invasive sub-retinal implantation of cells. Glass capillaries (OD: 1.2 mm, ID: 0.68 mm, WPI) were heated to 80°C with a puller, stretched, and ground with a grinder (Micro Grinder EG-400, Narishige, Tokyo, Japan) to a tip angle of 30°. The glass needle was then coated with either trichloro(1H,1H,2H,2H,2H-perfluorooctyl) silane (PFOCTS, Sigma Aldrich, Munich, Germany) or MPC (Tokyo Chemical Industry Co., Ltd, Tokyo, Japan). Thus, PFOCTS was placed in a sealed container with the glass needle and allowed to vapor-deposit for 24 h, whereas MPC was adjusted to 0.5 wt.% in ethanol and coated using aspiration and ejection with the glass needle. DPBS(-) and hyaluronic acid sodium salt (NaHA, Combi-Blocks, Inc., San Diego, CA, USA) adjusted to 0.5, 1.0, and 1.5 w/v% in DPBS(-) were used as a delivery solution for cell transplantation. A silicon tube was used to connect the glass needle to a syringe to create a delivery system. The syringe and silicon tube were filled with air or medical olive oil, and the cells on nanosheet were delivered at the tip of the needle.

In Vivo Assessment of Cells on Nanosheet Delivery to the Rat Retina

In this study, male Sprague-Dawley (SD) rats weighing 250-300 g (Japan SLC Inc., Hamamatsu, Japan) and Royal College of Surgeons (RCS) rats (CLEA Japan, Inc., Tokyo, Japan) aged 4–7 weeks were used. After receiving approval from the Tohoku University Environmental & Safety Committee’s Institutional Animal Care and Use Committee, all animals were handled in accordance with the Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Vision Research (no. 2014MdA-232-5). RPE-J cells on nanosheets were injected into the subretinal space using a hand-made glass capillary needle, after rats were anesthetized with ketamine hydrochloride (90 mg/kg) and xylazine hydrochloride (10 mg/kg). At an upper temporal lesion, we exposed the upper scleral surface and carefully punctured it with a 27 G gauge needle. Through the holes in the right eyes, we injected the cells on nanosheets into the subretinal space. Several days after implantation, the eyes were enucleated. For hematoxylin and eosin staining, eyes were fixed with 4% paraformaldehyde for 24 h. Standard paraffin-embedded sections (5 µm thick) were cut through the optic disk and stained with hematoxylin and eosin. Light microscopy images were taken. For cryosections, the enucleated eyes were frozen in OCT compound (Tissue-Tek; Sakura Finetek USA, Torrance, CA, USA) at -20°C and sectioning through the center of the implanted area at a thickness of 5 µm using a cryostat (Tissue-Tek Polar-D, Sakura Finetek. Ltd, Wetzlar, Germany).

Results

Mechanical Strength and Ease of Use of PCL/Collagen Nanosheet

To enable efficient mechanical handling, the material used to fabricate the nanosheet must have the appropriate rigidity and flexibility. Figure 3A showed the stress–strain diagram obtained from the tensile test used to assess the mechanical strength. The strain and maximum stress were high in PCL 100% (collagen 0%) and decreased with the decrease of the PCL concentration (PCL 50% collagen 50%). Furthermore, as the concentration of collagen increased, the maximum stress increased but the strain decreased. Next, we qualitatively investigated the mechanical properties of the nanosheets when they were immersed in water. As summarized in the table Fig. 3B, when the blending ratio of PCL/collagen was high, the nanosheet was fragile and broke easily, whereas when collagen was higher, the nanosheet had a great flexibility and could not maintain a flat shape, which made it difficult to use. PCL 80% /collagen 20% showed a moderate rigidity and flexibility and had the best operability, thus in the subsequent experiments, we used 80:20 ratio of PCL/collagen.

Figure 3.

Evaluation of mechanical strength and material property of the 20 mg/ml PCL/collagen composites. (A) Tensile stress-strain curve of a PCL/collagen test sample at varying mixture ratios. (B) Evaluation of the mechanical handling of the nanosheets with various PCL to collagen ratios. PCL: polycaprolactone.

PCL/Collagen Nanosheet Thickness

The PCL/collagen ratio being fixed, we evaluated the thickness of the PCL/collagen nanosheet by varying the PCL/collagen solution concentration from 5 to 100 mg/ml, and the spin coating rotation speed from 4,000 to 8,000 rpm for 40 s. Figure 4A shows that the thickness of the nanosheet was independent of the spin coating’s rotational speed in our condition of use. However, as shown Fig. 4B, the thickness of the nanosheet was determined by the concentration of the PCL/collagen solution, and as this concentration increases, the nanosheet thickens.

Figure 4.

Thickness of PCL/collagen nanosheet in function of (A) the spin coating speed and (B) the concentration of PCL/collagen. Values are the mean ± SD; n = 6. PCL: polycaprolactone.

Mechanical Strength and Biodegradability of Nanosheets

Nanosheets should have sufficient mechanical strength to remain in place during transplantation but should also be biodegradable with the tissue growth. We first compared the Young’s modulus of nanosheets made with 5 mg/ml PLGA, 20 mg/ml PCL, and 80:20 ratio of PCL/collagen at 20 mg/ml concentration. As shown Fig. 5A, with these different concentrations, when compared to PLGA nanosheet (84 MPa), PCL and PCL/collagen nanosheets had higher Young’s moduli (460 MPa and 293 MPa, respectively) and less flexibility under tension. Furthermore, when compared to PCL nanosheet, PCL/collagen nanosheet had a lower Young’s modulus and greater flexibility against tension due to the highly flexible physical characteristics of collagen ⁵⁰ . The mass loss of PLGA, PCL, and PCL/collagen nanosheets in the presence of 0.1 mg/ml lipase is shown Fig. 5B. It can be seen that the degradations of PCL and PCL/collagen nanosheets are milder than the one of PLGA nanosheet with 46% and 64% loss in mass at 200 days, compared to 77% loss in mass at 74 days, respectively.

Figure 5.

Evaluation of mechanical strength and biodegradability of nanosheets. (A) Young’s modulus of nanosheets made with 5 mg/ml PLGA, 20 mg/ml PCL, and 80:20 ratio of 20 mg/ml PCL/collagen. (B) Percentage of mass loss over time of nanosheets placed in lipase solution. Values are the mean ± SD; n = 4. PLGA: polylactic-co-glycolic acid; PCL: polycaprolactone.

Surface Wettability of Nanosheet

Figure 6A shows the appearance of water droplets on the different material surfaces. Additionally, Fig. 6B displays the calculated water contact angle. Thus, the water contact angles were 83.0°, 82.2°, 66.3°, and 55.3° for PLGA, PCL, PCL/collagen, and glass, respectively. Thus, PCL/collagen nanosheet was relatively hydrophilic.

Figure 6.

Contact angle of water on glass, PLGA, PCL, and PCL/collagen substrates. (A) Image of water droplets on various surfaces. (B) Contact angle measured for different surfaces. Values are the mean ± SD; n = 3. PLGA: polylactic-co-glycolic acid; PCL: polycaprolactone.

Cells Loaded Nanosheet

Figure 7(A–C) shows photos of cell-loaded nanosheet taken in brightfield and fluorescence microscopy. As shown, RPE-J cells were cultured confluently for 6 days on PCL/collagen nanosheet, stained with live/dead assay, and cell viability was high (Fig. 7C, D). Furthermore, 2 days after cell seeding, we divided the cell cultures into a group in which the cell-loaded nanosheet was detached from the glass substrate (free standing cell-loaded nanosheet) and a group in which it was not detached (schematic illustration). Cells on free standing nanosheet stayed perfectly fine with high viability (Fig. 7D) and proliferated normally (with 4.44 × 10³ and 5.05 × 10³ cells/mm² at days 4 and 6 of culture for the free-standing nanosheet), as cells on nanosheet attached to the glass substrate (with 2.24 × 10³, 4.26 × 10³, and 4.99 × 10³ cells/mm² at day 2, 4, and 6 of culture for the fixed nanosheet) (Fig. 7E).

Figure 7.

Photos of RPE-J cells stained with live/dead assay on PCL/collagen nanosheet taken in (A) bright field and (C) fluorescence microscopy. (B) Nanosheet stained with Nile red for better visualization (D) Cell viability on nanosheet over 6 days of culture with nanosheet attached to glass substrate (−) and free standing nanosheet (+). (E) Cell proliferation over time on nanosheet attached to glass substrate and on free standing nanosheet. RPE-J: retinal pigment epithelium-J; PCL: polycaprolactone.

Cell Morphology

A Voronoi diagram is a method of geometrically dividing a plane according to the nearest neighbor rule at points on the plane ⁵¹ . Epithelial cells such as RPE-J can be evaluated for their morphology using Voronoi diagrams ⁵² . RPE-J cells were stained for tight junction with anti-ZO-1 antibodies and visualized in red fluorescence with a secondary dye-conjugated antibody. Figure 8A shows tight junction of cells. In Fig. 8B, the Voronoi diagram was created from the center of gravity of the nucleus obtained from the fluorescence intensity in Fig. 8A using image processing software. The cell contour was estimated with a Voronoi diagram from the position of the nucleus and RPE-J cells showed a pavement-like morphology. Figure 8C shows the cell repartition in polygonality from the cell layer analysis. Thus, 42% and 47% of cells were hexagonal at day 4 and day 6 of culture on the free-standing nanosheets, compared to 38% and 42% on the fixed nanosheets.

Figure 8.

Cell morphology evaluation. (A) Photo in fluorescence microscopy of RPE-J cells at day 6 of culture stained against ZO-1 protein ¹³ , and cell nuclei (blue). Tight junctions are perfectly visible, and therefore, we may think that the barrier function of RPE-J cells has been established. (B) Voronoi diagram for an RPE-J cell layer. (C) Distribution of RPE-J cells in polygonality. RPE-J: retinal pigment epithelium-J.

Cell Loaded Nanosheet Delivery Evaluation

Figure 9(A, B) shows a comparison of the fabricated glass needle we used to deliver the cell-loaded nanosheet and a 27 G (220 µm inner diameter) medical injection needle. The fabricated glass needles had an inner diameter of 0.22 ± 0.03 mm and an outer diameter of 0.41 ± 0.02 mm. The thinner the needle used for transplantation is, the less invasive it is. However, if the needle is too thin, it will be difficult to aspirate and release the cell-loaded nanosheet, and there is concern that cells will be damaged by friction. Thus, we considered that a needle with a thickness similar to the one of a 27-gauge needle was suitable for a reliable cell-loaded nanosheet delivery whereas suppressing invasiveness. Figure 9C shows a schematic diagram of the implantation site of the cell-loaded nanosheet. The glass needle was inserted through the sclera from the posterior segment of the eye to the sub-retinal space. Figure 9D is a photo taken in merged brightfield and fluorescence microscopy which shows the cell-loaded nanosheet in the glass needle.

Figure 9.

Overview of the cell delivery system used. (A, B) Size comparison of the glass needle used in our experiments and a stainless steel 27-gauge needle. (C) Schematic showing the insertion site of the glass needle for cell loaded nanosheet delivery to the sub-retinal space. (D) Photo in merged bright field and fluorescence microscopy showing a cell loaded nanosheet (red-green fluorescence) in the glass needle. (E) Photo of the system used to deliver the nanosheet. (F) Evaluation of the nanosheet delivery with coated and non-coated glass needles. (G) Evaluation of the nanosheet delivery loaded in solutions with different viscosities (× negative, Δ bad, O good, ◎ excellent). (H) Evaluation of the delivery of nanosheet in solution with the use of air pressure and oil pressure. PFOCTS: Trichloro(1H,1H,2H,2H-perfluorooctyl)silane; MPC: 2-methacryloyloxyethyl phosphorylcholine; PBS: phosphate-buffered saline.

An overview of the cell-loaded nanosheet delivery system we used is shown in Fig. 9E. It consists of a 1-ml syringe and a glass needle connected with a silicon tube. We evaluated the delivery rate of nanosheets from the glass needles with or without coating. Thus, using PFOCTS-coated and MPC-coated glass needles, cell loaded nanosheets were loaded and released from the glass needles, and the results on the rate of nanosheet remaining in the needle are shown Fig. 9F. Thus, when glass needles without coating were used, the delivery of cell-loaded nanosheet was poor with a rate of nanosheet remaining in the glass needles of 48%. When glass needles coated with PFOCTS were used, the delivery of cell-loaded nanosheet improved a little and the rate of nanosheet remaining in the glass needles decreased to 38%. The best cell-loaded nanosheet delivery was obtained with glass needles coated with MPC, and the rate of nanosheet remaining in the glass needles was only 20%. Therefore, in the next experiments, we used MPC-coated glass needles.

Figure 9G shows the results of nanosheet delivery when compared different delivery solutions such as DPBS(-) used as a control, and different concentrations of sodium hyaluronate (NaHA) in DPBS(-). We achieved the best nanosheet delivery with NaHA 1% in DPBS(-), while the solution viscosity was moderated and allowed good loading of the nanosheet into the syringe. Therefore, we decided to use NaHA 1% in DPBS(-) for cell-loaded nanosheet transplantation in further experiments. Figure 9H shows the successful delivery rate of cell-loaded nanosheet to the sub-retinal region of rat eyeball under air pressure (27%), and oil pressure (83%). In the case of air pressure, the solution with the cell-loaded nanosheet came out from the syringe with difficulties because the tip of the needle was blocked by the eyeball tissue, and when the syringe plunger was pushed more strongly, the pressure reached a certain level and the solution was ejected too quickly together with the air, which resulted in difficulties to deliver the cell-loaded nanosheet to the target site. On the other hand, in the case of oil pressure, it was possible to deliver the cell-loaded nanosheet while finely adjusting the release flow rate by operating the syringe.

Cell Loaded Nanosheet Delivery in Vivo

Figure 10A shows the delivery test of the cell loaded nanosheet to Sprague-Dawley (SD) rat eyes. Figure 10(B–D) shows the appearance of the eyeball removed 2 days after transplantation. The PCL/collagen nanosheet was stained with Nile red to help in its localization, and RPE-J cells were stained with calcein-AM, both were confirmed under fluorescent light. The following photos are photos in brightfield (10E), fluorescence microscopy (10F, G) and merged (10H) of a cross section of the eyeball after staining. The fluorescence of the cell-loaded nanosheet was confirmed from the sub-retinal area, which proved a perfect delivery to the target site. Furthermore, PCL/collagen nanosheet and PLGA nanosheet were implanted in rat eyes, and the inflammation levels at day 3 and day 8 post-surgery were compared. Figure 10(I, J) shows the cross-section of the fundus tissue after hematoxylin and eosin staining of the paraffin section of the removed eyeball. A space has been formed between the RPE cell layer and the photoreceptor layer, due to the delivery of the RPE-J cell-loaded nanosheet in solution. Concerning this space a natural healing is expected. In Fig. 10I, no immune cells infiltration was observed, indicating that PCL/collagen nanosheet did not trigger an inflammation reaction. Interestingly, immune cell infiltration was also not observed with PLGA nanosheet (Fig. 10J).

Figure 10.

Cell-loaded nanosheet transplantation in vivo. (A) Photo of eye surgery on rat. (B-D) Photos in bright field and in fluorescence microscopy showing the treated eyeball 2 days post-operation with the cell loaded PCL/collagen nanosheet. Photos in bright field (E) and in fluorescence microscopy (F, G) and merged (H) showing a cross section of the eyeball with the cell loaded PCL/collagen nanosheet delivered to the sub-retinal space. (I, J) Photos of eye cross sections stained with hematoxylin and eosin showing no immune cell infiltration after 2 days post-transplantation of cell-loaded nanosheets made of (I) PCL/collagen and (J) PLGA. (K, L) Photos of eye cross sections stained with hematoxylin and eosin showing thick photoreceptor layer after 2 weeks post-transplantation of (K) cell-loaded PCL/collagen nanosheet, compared to (L) PCL/collagen nanosheet without cells used as a control. Scale bars: 2 mm (B-D), and 200 µm (E-L). PCL: polycaprolactone; PLGA: polylactic-co-glycolic acid.

RCS rats show abnormal phagocytosis of photoreceptor outer segments by retinal pigment epithelial cells. After birth, the retina forms normally, but photoreceptor degeneration begins around 3 weeks after birth, and most photoreceptors disappear around 3 months after birth ⁵³ . Cell-loaded PCL/collagen nanosheet and PCL/collagen nanosheet without cells (control) were implanted in RCS rat eyes and removed after 2 weeks. Figure 10(K, L) shows the cross-section of the fundus tissue after hematoxylin and eosin staining of the paraffin section of the removed eyeball. It can be seen that RPE-J cells formed lumps of cells in the subretinal space and that the thickness of the photoreceptors layer at the transplanted site has increased when the RPE-J cell-loaded nanosheet was transplanted (K), compared to the control (L). Thus, the extracellular environment was rebuilt by the barrier function of the transplanted RPE-J cells, and secreted factors have been released preventing the disappearance of photoreceptors.

Discussion

Delivery of RPE cells to the sub-retinal space is an important strategy against ophthalmic disease such as AMD, to restore RPE cells homeostasis and functionality, provided the transplanted cells can generate a monolayer, regenerate the RPE/photoreceptors interface, and inhibit choroidal vessel invasion ⁵⁴ . The use of polymeric nanosheet as a cell carrier is a very interesting approach which is well adapted to the delivery of organized cells in such narrow space and delicate environment ²⁷ . For the fabrication of PCL/collagen nanosheet, our first goal was to optimize the PCL/collagen ratio to meet the mechanical strength and ease of handling requirements. The evaluation of the mechanical strength with different ratios of PCL/collagen showed a decrease in strain with the increase in collagen, which is consistent with results obtained in other studies ⁵⁰ . Assuming that the binding between PCL-PCL polymeric chains and collagen–collagen polymeric chains are stronger than between PCL-collagen chains, the resulting binding force would be dominated by the material that is dominant in the bulk sample. Furthermore, we evaluated qualitatively the mechanical properties (handling) of the nanosheet in water and selected the PCL/collagen 80:20% ratio for its best operability due to its moderate rigidity and flexibility. Next, we evaluated the thickness of the PCL/collagen nanosheet in function of the spin coating rotation speed and observed no significative variation in thickness. This is due to the fact that the amount of polymeric solution left on the stamp remains constant regardless of the rotation speed as long as a sufficient rotation time of 40 s was ensured. Since the amount of the PCL/collagen solution remaining on the stamp does not change after 40 s spin coating, the thickness of the nanosheet is then determined by how viscous the material is, and the material’s viscosity increases with its concentration. Therefore, changing the concentration of the polymer solution can be used to tune the thickness of the nanosheet. As a thickness around 300 nm works best as substrate for cells, we chose the concentrations of 16 mg/ml PCL and 4 mg/ml collagen to fabricate the nanosheet. Then, we assessed the Young’s modulus of the PCL/collagen (80:20%) nanosheet. As expected, it was lower than the Young’s modulus of PCL nanosheet with similar concentration (20 mg/ml) due to the presence of collagen which has higher flexible property than PCL ⁵⁰ . When compared with PLGA nanosheet, the Young’s modulus of the PCL/collagen nanosheet was higher, showing less flexibility, but the concentration of PLGA was different (5 mg/ml), which limits the comparison. The biodegradability of polymeric scaffolds is also an important parameter in tissue engineering as in cell delivery. As expected, both PCL and PCL/collagen nanosheets degraded more slowly than the PLGA nanosheet in presence of lipase. As mentioned previously, such property could be useful if a sustained controlled drug delivery is also planned with the cell delivery to support the cell engraftment. PCL degrades by hydrolysis of poly(-hydroxy) esters via surface or bulk degradation pathways ⁴⁵ . If in PCL and PCL/collagen nanosheets the rate of PCL hydrolysis and by-product diffusion is faster than the rate of water infiltration into the bulk polymer, the degradation pathway is a surface degradation ⁵⁵ . Internal catalysis, on the other hand, accelerates the polymer degradation in the bulk degradation pathway, rapidly releasing acid by-products ⁵⁶ . Since the thicknesses of the nanosheets are around 300 nm, we assume that the degradation of the PCL and PCL/collagen nanosheets is occurring via a surface pathway. Next, we evaluated the hydrophilicity of the different nanosheets since the cell adhesion and spreading are higher on hydrophilic surfaces ⁵⁷ . Thus, the cell adhesion rate is greatest when the contact angle of water is around 60°–70° ⁵⁸ ; therefore, the wettability of PCL/collagen nanosheet suggests an excellent substrate for cell adhesiveness.

After this material part characterization, we focused on the biological part and began by evaluate the cell viability and proliferation. The results obtained showed that RPE-J cells attached well on PCL/collagen nanosheet, have an excellent viability, and proliferated normally. It is well known that PCL is a good substrate for cells but has some hydrophobicity; therefore, the presence of collagen, which has strong cell supportive properties, enhances the cytocompatibility of PCL ⁴⁷ . Interestingly, we also observed that RPE-J cells cultured on a free-standing PCL/collagen nanosheet did not differ in terms of cell viability and proliferation than RPE-J cells cultured on fixed PCL/collagen nanosheet. This is important because we want to transplant the cells on the nanosheet (free-standing); therefore, we need them in perfect condition. Moreover, we analyzed the RPE-J cell layer formed with a Voronoi diagram based on the nearest neighbor rule. The nearest-neighbor rule is a method of classifying a plane region closest to each point. In other words, it can be obtained by drawing the perpendicular bisector between adjacent generating points and dividing the nearest neighbor region of each generating point. It is known that Voronoi diagrams occur in various situations in the natural world, and various researches have been conducted to capture natural phenomena by creating Voronoi diagrams^59,60. Our analysis showed that RPE-J cells formed a densely packed structure with a cobblestone-like morphology. The hexagonal shape is known to be the most stable configuration of cells allowing the greatest surface coverage with minimal surface energy^49,61. In normal individuals, the RPE cell density near the macula is around 4,964 cells/mm², and the ratio of hexagonal RPE cells is around 40.17% ⁶² , which is in agreement with the results we obtained with RPE-J cells in culture (Fig. 8C). Therefore, the PCL/collagen nanosheet is an excellent substrate that supports RPE-J cells in culture, and their natural formation in a cellular layer.

Then, we focused on the cell delivery process. We used a very thin glass needle (thickness similar to a 27 G needle) coupled to a silicon tube connected to a 1 ml syringe, and injected the cell loaded PCL/collagen nanosheet through the sclera from the posterior segment of the eye to the sub-retinal space. This way enables the delivery of cell loaded nanosheet without damaging the intraocular or retina. In addition, by setting the tip angle of the needle to 30°, it became possible to insert the needle along the retina, making the operation of transplantation easier to perform. However, before setting a real operation on rats, we assessed the delivery of the nanosheet with our system and optimized it. It appeared that the delivery rate of the nanosheet using the glass needle without coating was poor (52%). Thus, we coated the glass needle with PFOCTS which is a chlorosilane that forms self-assembled monolayers on various substrates, and is mainly used for superhydrophobic surface modification and anti-adhesion ⁶³ . This treatment improved the nanosheet delivery rate (62%). However, the best cell-loaded nanosheet delivery rate (80%) was obtained with glass needles coated with MPC, which has a phosphorylcholine group in its side chain, mimicking the polar group of phospholipids contained in cell membranes ⁶⁴ . This greatly improved the lubricity and hydrophilicity of the glass needle ⁶⁵ . Next, we evaluated the injection process. We tested different solutions (DPBS(-), NaHA at different concentrations) to be injected with the cell loaded PCL/collagen nanosheet. We observed that the loading and releasing of the nanosheet was improved when the solution has some consistency or viscosity (which eliminates the DPBS(-)). However, if the viscosity of the solution is too high, it will not be possible to aspirate it or release it with the glass needle. Thus, the best solution tested for nanosheet delivery was NaHA 1% in DPBS(-). Moreover, we also observed that the nanosheet delivery was difficult with pneumatic (air) pressure when the glass needle was in biological tissue, due to the interstitial pressure of the tissue ³¹ . To overcome this problem, we used oil (olive oil for medical use) pressure which is not compressible compared to air; therefore, the movement of the syringe plunger and the movement of the solution containing the cell-loaded nanosheet at the tip of the needle were interlocked without any loss.

Finally, we made in vivo experiment on Sprague Dawley rat eyes. The delivery of the cell-loaded nanosheet to the sub-retinal space was a success validating our methods of injection. Furthermore, histological analysis was made in days 3 and 8 post-operation to compare the immune response activation when RPE-J cell-loaded PCL/collagen and RPE-J cell-loaded PLGA nanosheets were transplanted. The results showed no immune response activation in both cases. Although we expected no (or low) immune reaction with PCL/collagen nanosheet, these results were less expected concerning PLGA nanosheet because the by-products generated in vivo by PLGA degradation are known to trigger inflammation ⁶⁶ . Therefore, it is possible that because the PLGA nanosheet is very small, 300 nm thick and 200 μm in diameter, the amounts of by-products generated by the PLGA degradation was very small and below a threshold that trigger an inflammation reaction. Importantly, we transplanted in RCS rat’s eyes PCL/collagen nanosheets with RPE-J cells and without (control), and performed histological analysis 2 weeks later. We observed that RPE-J cells formed lumps of cells in the sub-retinal space, whereas the thickness of the photoreceptor layer at the transplanted site has increased. This is thought to be because the extracellular environment was reconstructed by the barrier function of the transplanted RPE-J cells and the released factors preventing the disappearance of the photoreceptors. Therefore, the delivery of RPE-J cell loaded PCL/collagen nanosheet is an efficient strategy in treating ocular diseases by substituting for host RPE cells and by restoring the retinal environment.

In conclusion, in this study we fabricated a composite PCL/collagen nanosheet used as a cell carrier to deliver RPE-J cells to the sub-retinal space in a minimally invasive manner, and compared it to a PLGA nanosheet. Indeed, both materials are biodegradable, however PCL has several advantages over PLGA. Thus, it degrades in vivo more slowly than PLGA, which is suitable for long-term controlled drug and growth factor delivery to improve cell functionalities and cell integration to the host. Furthermore, the by-products of its degradation are less likely to induce an immune reaction. Moreover, for a given concentration, PCL-based nanosheets showed usually better flexibility than PLGA nanosheets, which gives to cells better protection during injection and potentially allows the use of needles with thinner diameter. In addition, the presence of collagen promotes cell attachment, proliferation, and functions.

In our study, the PCL/collagen nanosheet showed tunable mechanical properties, moderate rigidity and flexibility enabling manipulation, and was an excellent substrate for RPE-J cell culture since the cell viability and proliferation were high, and the cells formed quickly an epithelial layer with tight junctions. Furthermore, we optimized the delivery of the cell loaded nanosheet in solution using a customized system based on a glass needle connected to a syringe with a tubing. Experiences in vivo on RCS rat eyes showed that the RPE-J cell-loaded nanosheet was efficiently delivered to the sub-retinal space, and confirmed its therapeutic effect by substituting for host RPE cells and regenerating the retinal environment. Thus, the method of delivery and the PCL/collagen nanosheet presented in this study have great potential in the transplantation and local delivery of organized cells, and can be useful to the treatment of other retinal diseases.